Affective Polarization, Moralized Values, and Political Sectarianism in Contemporary Democracies

Polarização afetiva é a hostilidade emocional ao partido oposto — fenômeno distinto da mera divergência programática. O texto mapeia como identidade de grupo e moralização do conflito se combinam no que Eli Finkel e coautores chamam de “sectarismo político”: uma configuração em que opositores são vistos não como errados, mas como moralmente contaminados, tornando o compromisso sinônimo de traição e dificultando soluções procedimentais.

Para os interesses do vault, o conceito é chave para entender o bolsonarismo e a corrosão das normas democráticas no Brasil e alhures. A literatura documentada mostra que polarização afetiva se mede em escalas multidimensionais — othering, aversion, moralization —, com efeitos distintos sobre atitudes antidemocráticas e apoio à violência política. Intervenções que corrigem metapercepções exageradas podem reduzir desumanização, abrindo uma agenda prática.

A literatura associada a Iyengar, Mason, Skitka e Van Bavel demonstra que hostilidade partidária cresce independentemente de divergência ideológica; que “mega-identidades” (quando raça, religião, educação e partido se alinham) intensificam o viés mesmo com posições de política constantes; e que mídia e redes sociais amplificam moralização via incentivos de engajamento. Resultados comparativos em 40 democracias (CSES) indicam que a polarização afetiva varia com instituições e liderança — não é destino.

Scope, definitions, and the central hypothesis

This research tests a specific interpretive hypothesis: that contemporary political polarization is driven less by programmatic (policy/ideological) disagreement and more by group identification plus moralization, such that political differences are reinterpreted as fundamental moral differences—fueling dehumanization, legitimizing “anything goes” tactics, and weakening democratic norms.

Three distinctions anchor the inquiry.

Affective polarization refers to polarized feelings—warmth toward one’s own side and hostility toward the other side—rather than polarized issue positions. In practice, it is commonly operationalized as gaps in evaluative ratings (e.g., “feeling thermometers”), social-distance items (e.g., discomfort with an out-party marriage), trait attributions, and behavioral discrimination.

Ideological polarization refers to divergence in policy preferences, ideological self-placement, and (often more strongly) elite positions. A recurring theme in the literature is that intensifying hostility can appear disproportionate to measured policy divergence—especially among mass publics—suggesting at least partial decoupling between policy disagreement and social enmity.

Moralization (moralizing a political value) is not merely “caring strongly.” It means experiencing an attitude as rooted in core beliefs about right and wrong (“moral conviction” / “moral mandate”), which tends to make disagreement feel illegitimate, compromise feel like betrayal, and procedural solutions harder to generate.

A key synthesis proposal—central to testing the hypothesis—is that what looks like a single phenomenon (“polarization”) is often better described as a bundle of othering, aversion, and moralization operating together, which Eli J. Finkel and coauthors call political sectarianism.

Freeden-style conceptual morphology of polarized politics

The required analytic framework comes from Michael Freeden’s morphological approach: political meaning is organized as configurations of concepts, whose relative weights and proximities shape how political language is “decontested” (stabilized enough to guide action).

A practical way to apply this to contemporary polarization is to treat polarized political conflict as a conceptual pattern with (a) core concepts (dominant and organizing), (b) adjacent concepts (interpreting/supporting the core), and (c) peripheral concepts (more contingent, policy-specific, operational). A clear summary of this core/adjacent/peripheral logic—and of “priority” and “proximity” as key morphological features—is provided in an open-access methodological discussion of Freeden’s approach.

Two competing morphologies help clarify what is new (or newly intensified) in contemporary democracies.

A pluralist-democratic morphology (competitive but bounded conflict):

• Core: democracy (rules, rights, legitimacy), pluralism (many parties/values), cooperation/compromise as legitimate, institutional trust as conditionally earned.
• Adjacent: disagreement, accountability, representation, opposition as loyal rival.
• Peripheral: specific policy bundles; leaders; episodic scandals; shifting coalitions.

This configuration treats the opponent as wrong but legitimate, and it keeps policy disagreement in the same moral universe as bargaining and revision.

A sectarian-moralized morphology (antagonistic “us vs them”):

• Core: group identity (“who we are”), moralization (“we are the good”), enemy/traitor schema (“who they are”), conflict as existential.
• Adjacent: distrust, perceived threat, humiliation, belonging, purity/contamination narratives, loyalty signals, punishment of deviants/compromisers.
• Peripheral: concrete policy details, empirical uncertainty, tradeoffs, proceduralism, and sometimes even factual accuracy.

This configuration explains why the same policy disagreement can be processed either as negotiable conflict (pluralist morphology) or as a moral emergency (sectarian morphology). It also predicts why “compromise” can morph from core democratic practice into a stigmatized peripheral act.

Crucially, the Freeden lens makes the hypothesis testable in conceptual terms: if contemporary polarization is disproportionately driven by identity and moralization, then we should observe (i) identity and morality concepts moving toward the core, (ii) policy specifics moving toward the periphery, and (iii) the “opponent” concept shifting semantically from adversary to enemy/evil. Multiple empirical literatures align with that pattern.

Affective polarization and measurement

The modern literature on affective polarization—strongly associated with Iyengar’s work—frames polarization as social distance and evaluative hostility, and explicitly argues that focusing only on issue extremity misses a central dimension of contemporary conflict.

Core empirical measures

Common measurement strategies include:

Feeling-thermometer gaps (e.g., 0–100 warmth ratings). A major integrative discussion reports decades-long decline in out-party warmth, while in-party warmth is comparatively stable—consistent with the idea that “out-party hate” has become more defining than “in-party love.”

Social distance items (marriage, friendship, neighborhood comfort). These connect partisan hostility to everyday life rather than to abstract ideology, capturing how political division becomes social segregation.

Behavioral discrimination experiments (e.g., hiring, allocation, trust games), which test whether partisan labels drive real choices, not just survey talk.

Measurement pitfalls and a recent methodological upgrade

Two problems are increasingly emphasized:

1. Some “affective polarization” indices may partially capture attitudes toward elites/leaders rather than ordinary voters.
2. Treating affective polarization as one-dimensional can obscure distinct psychological components.

A major 2026 contribution in American Political Science Review directly addresses both by proposing and validating a multidimensional scale with three subdimensions: othering, aversion, and moralization. It also reports that these components have different correlational patterns with outcomes like anti-democratic attitudes and support for political violence—exactly the kind of nuance needed to test the hypothesis (moralization is not just “more dislike”).

Taken together, the measurement literature supports a key part of the hypothesis: the most politically relevant polarization for democratic stability is often affective and moralized, not merely ideological—but it also warns that we should not equate every thermometer gap with the same underlying mechanism.

Partisan identity as social identity

Affective polarization becomes politically dangerous when “party” behaves like a social identity (with in-group/out-group psychology), rather than merely an instrument for choosing policies. The social-identity view emphasizes categorization, in-group favoritism, out-group derogation, identity signaling, and threat sensitivity.

Two mechanisms repeatedly show up as drivers.

Social sorting into “mega-identities.” When ideology, geography, religion, education, race, or culture align more cleanly with party coalitions, partisan identity becomes thicker—less about policy preference and more about “people like us.” Lilliana Mason’s work is foundational here: sorting intensifies bias, anger, and activism even when issue positions are held constant, suggesting a route from identity alignment to social polarization.

Negative partisanship (“out-party hate”). Evidence shows that hostility toward the opposing party can operate independently of strong positive attachment to one’s own party, and that it can strongly predict political behavior.

The identity account directly supports the central hypothesis, but it also introduces an important limit: identity dynamics do not require the absence of real ideological disagreement; rather, they can amplify and reinterpret disagreement. That matters because some experimental evidence indicates that perceived ideological divergence can increase hostile affect, implying partial policy→affect pathways that the strongest form of the hypothesis might understate.

Moralization of values and motivated reasoning

Moralization is where disagreement crosses a qualitative threshold: opponents are not merely mistaken—they are experienced as immoral.

Moral conviction as a political amplifier

Linda J. Skitka’s synthesis of decades of work treats moral conviction as a distinct psychological dimension with predictable downstream effects: stronger emotions, less compromise, more intolerance toward dissenters, and greater readiness to endorse extreme means when moral ends feel threatened.

In parallel, Timothy J. Ryan’s empirical work on “no compromise” finds that moralized attitudes lead citizens to oppose compromise and punish compromising politicians—even when compromise would produce material gains—helping explain why negotiated solutions collapse once issues become moral mandates.

There is also evidence that moralization can be socially and emotionally induced: exposure to anger- and disgust-eliciting frames can moralize attitudes and increase polarization, plausibly linking media rhetoric to value absolutism.

What the Haidt and Kahan lines add—and what they don’t

Contrast frameworks help identify the boundaries of the central hypothesis.

In The Righteous Mind and in the broader “social intuitionist” tradition, the basic claim is that moral judgments often arise from fast intuitions, with reasoning frequently recruited post hoc. The core theoretical statement is clearly articulated in a widely circulated version of the 2001 “emotional dog” model. This supports the plausibility of “moralization first, justification later” dynamics in politics.

In the cultural cognition tradition, Dan Kahan’s work argues (with experimental and correlational evidence) that people’s perceptions of contested facts (e.g., scientific consensus) are shaped by identity-congruent meanings and group commitments (“cultural cognition”), not just information deficits. This strengthens the identity-centered mechanism: factual disputes can become identity disputes, which can become moral disputes.

At the same time, these frameworks should not be treated as universal “polarization engines.” For misinformation specifically, evidence suggests that inattention and weak analytic engagement can be more predictive than pure identity-motivated bias, and that simple accuracy prompts can reduce misinformation sharing. That creates a real boundary condition: not every “bad belief” is best explained by motivated reasoning; some is explained by low attention, platform incentives, and cognitive shortcuts.

A useful integration is the identity-based model proposed by Jay J. Van Bavel and coauthor: partisan identity can serve multiple goals (belonging, status, moral meaning), and those goals can outweigh accuracy goals—so bias is contingent on goal tradeoffs, not inevitable.

Dehumanization, demonization, and affective escalation

If moralization is the “qualitative jump,” dehumanization is the “norm-breaking jump”: it places opponents outside normal moral consideration.

What dehumanization is, and how it is measured

Dehumanization is not one thing. A classic integrative review distinguishes “animalistic” and “mechanistic” forms, emphasizing that denying distinct forms of humanness has different psychological bases and can occur outside extreme violence contexts.

A major empirical innovation is the “Ascent of Man” measure, designed to capture blatant dehumanization (explicitly rating groups as less evolved/less human). This measure is validated and widely used across intergroup contexts.

A second key construct is metadehumanization—believing the outgroup sees your group as less than human—which predicts reciprocal dehumanization and aggressive policy support.

Why dehumanization matters for democratic life

Dehumanization is a plausible microfoundation for democratic norm erosion because it weakens the intuitive brakes on cruelty, exclusion, and rights violations. Empirically, exaggerated meta-perceptions—overestimating how prejudiced or dehumanizing the other side is—predict hostility, implying that polarization can be self-reinforcing through misperceived moral threat.

There is also credible evidence that interventions correcting exaggerated meta-perceptions can reduce explicit dehumanization, suggesting at least one actionable pathway that targets the “enemy” concept directly (a core node in sectarian morphology).

Finally, these dynamics are not unique to the U.S.-centered literature: recent comparative work examines partisan dehumanization in Brazil’s asymmetrically polarized party system, illustrating transferability of dehumanization frameworks to different institutional ecologies.

Media, social networks, misinformation, and escalation dynamics

The media-and-platforms question is not “do people live in total echo chambers?” but “what aspects of the information environment reliably intensify identity conflict and moralization?”

Moral-emotional virality and “outrage incentives”

Large-scale observational work finds that moral-emotional language increases diffusion of political content on social networks, consistent with the claim that platforms—via engagement incentives—can reward moralized expression and intensify within-group reinforcement.

Echo chambers are real but partial—and sometimes the backfire is worse

Evidence from large platform datasets suggests that both network structure and individual choices limit cross-cutting exposure, even when diverse content is available.

At the same time, “just show people the other side” is not a reliable depolarization strategy. A large field experiment finds that exposure to opposing political views on social media can sometimes increase polarization, with asymmetric effects across groups.

A balanced reading is that platform effects are conditional: they depend on selection, identity threat, and the emotional framing of content—not merely on whether cross-cutting content exists.

Misinformation and the escalation spiral

False news spreads faster and further than true news on Twitter in a large-scale analysis of verified stories, a pattern plausibly relevant to polarizing dynamics because novelty and emotion can be mobilized for identity and moral conflict.

But the best evidence also suggests that scalable interventions can reduce harm: shifting attention to accuracy reduces misinformation sharing, including in platform field settings. This matters because misinformation is a common accelerant in sectarian escalation—especially when framed as moral betrayal or existential threat.

Institutions, international variation, and implications for democracy

The “inevitable vs contingent” question is best answered by combining cross-national evidence with democratic theory.

Cross-national variation implies contingency, not destiny

Cross-national work using Comparative Study of Electoral Systems data compares affective polarization toward parties and leaders across 40 democracies over more than two decades, showing that affective polarization is widespread but varies substantially across countries and election contexts—suggesting that institutions, party systems, and leadership styles shape outcomes.

This finding aligns with the broader comparative argument that affective polarization is not a uniquely American pathology—even if certain U.S. dynamics are especially intense.

Democratic consequences are most acute when polarization becomes “pernicious”

A strong comparative framework differentiates ordinary polarization from “pernicious polarization,” where societies split into mutually distrustful “us vs them” camps in ways that damage democracy. This framework emphasizes agency (leaders and parties) and interaction between demand-side grievances and supply-side elite strategies.

On the mass-public side, experimental and observational evidence indicates that polarization can weaken the public’s willingness to punish democratic norm violations when committed by co-partisans, undermining voters as a democratic check.

At the same time, the newest measurement work warns against overclaiming: evidence linking affective polarization to anti-democratic attitudes is strong in some studies but not uniform across operationalizations, reinforcing the need to distinguish subdimensions such as othering vs moralization vs aversion.

The inevitability debate through pluralist, agonistic, and “friend–enemy” lenses

From pluralist democratic theory (e.g., Robert A. dahl’s account of democracy’s institutional requirements), conflict among groups is normal; democracy is partly a system for managing disagreement without domination.

From value pluralism (e.g., Isaiah Berlin), deep value conflicts can be ineliminable; the problem is not disagreement itself but the belief that disagreement must be resolved by eliminating the other.

From agonistic democratic theory (e.g., Chantal Mouffe), the democratic task is transforming antagonism (enemies) into agonism (legitimate adversaries).

And from Schmittian critique (summarized in a high-quality reference work), “the political” can be conceived as grounded in friend–enemy distinctions—an arc that mirrors what happens when sectarian morphology becomes dominant.

Across these lenses, the most defensible conclusion is: some polarization is a feature of mass democracy and pluralism, but contemporary affective–moralized polarization is not inevitable. It is contingent on social sorting, media incentive structures, elite strategies, and institutional designs that heighten the payoff of zero-sum conflict and identity-based mobilization.

What follows for solutions, and their limits

No single “depolarization hack” is credible. The evidence points to targeted levers, each with constraints:

Psychological levers that target misperceptions and dehumanization can work, but typically yield modest effects and require scale. Information-quality interventions (accuracy prompts) can reduce misinformation sharing, but they do not solve identity conflict by themselves. Media exposure to opposing views can backfire under identity threat, so “more cross-cutting contact” must be designed carefully (tone, context, trust, incentives). Institutional and elite incentives matter because they can either reward moralized demonization or impose costs on norm-breaking and anti-democratic tactics.

A Freeden-style restatement of the solution problem is blunt: reducing democratic risk requires reweighting the conceptual morphology of politics—pushing “democracy,” “legitimacy,” and “cooperation” back toward the core, while pushing “enemy,” “moral contamination,” and identity absolutism away from the core. That is possible, but not automatic—and it depends as much on institutions and elites as on individual psychology.

Ver também

• culturalcognition — a cognição cultural de Kahan é um mecanismo paralelo: percepções de fatos são moldadas por identidade de grupo, não apenas por déficit informacional; os dois frameworks se complementam para explicar por que corrigir informação não desfaz a polarização
• lakoff_haidt_kahan — Haidt e o modelo intuicionista social explicam por que moralização precede justificação: conexão direta com a seção sobre moral conviction como amplificador político
• democraticerosion — polarização afetiva é o mecanismo de demanda que facilita erosão democrática: o público co-partisan tolera norma-breaking do próprio lado quando a hostilidade ao outro partido é alta
• thymos — thymos e polarização afetiva são mecanismos distintos mas se reforçam: a aversion e o othering alimentam megalothymia (necessidade de superioridade) e negam isothymia (reconhecimento igual) ao adversário
• dahl — o texto usa dahl como contraponto teórico: pluralismo democrático vs morfologia sectariana são dois modos de organizar o conflito político legítimo